2020 was a big year for Outer Join–our first! To commemorate the year of our launch and document the jobs that we found along the way, I’d like to share a bit of company history and explore the inner world of data at Outer Join.
I’m a software developer by trade and in the spring of 2020 indulged a long-running curiosity about data science: I taught myself some fundamentals (NumPy, Pandas, Scikit-learn, etc) while also volunteering with my daughter's school system to help them better leverage data. I was having fun and started looking for a data analyst role that would fit my lifestyle (read: remote).
While exploring the job market, I struggled to find a high-quality, niche job board for data science like those I’d frequented in programming and design. Having a long history and interest in hiring platforms from my experience co-founding and running Dribbble, this felt like an opportunity.
Coding ensued, a wordplayful domain was secured, and a site was born. To prepare it for public consumption, I got help from one of the best designers in the business–Dan Cederholm–who’s responsible for the double-entendred logo and heroic splash image that regales the site. I also engaged designer/developer extraordinaire Andy Leverenz, who helps keep the UI looking sharp, well-maintained, and working across devices.
Outer Join was live (but not linked) for much of the summer of 2020 as I grew the corpus of jobs and hiring organizations. It was unveiled when I reached out to the folks at Data Science Weekly in September 2020, who were kind enough to provide a mention in their newsletter which resulted in our first visitors. Blast off!
I was thrilled to see that mention followed shortly by another in The Data Science Roundup curated by Tristan Handy, CEO & Founder of Fishtown Analytics, makers of dbt. He had this to say about Outer Join:
!! Super-cool. Lots of opportunities posted on there at high-quality companies. From what I can tell, this is quite new.
Whoa! Super-cool indeed. Thanks, Tristan!
There's much more outreach planned for 2021, but the remainder of 2020 was spent mostly heads down, working on infrastructure to import the latest jobs and hiring organizations and laying the groundwork to build out the platform and grow in 2021.
But enough about our story–I’d rather shine a light on the jobs and hiring organizations discovered along the way. So let’s look at some data ...
I launched the site with three core categories into which jobs were classified:
Before long, it became clear that a non-trivial swath of roles wasn’t quite fitting any of these and warranted a category of its own, so I created a fourth (and reclassified existing jobs):
Machine Learning (ML) Engineering
Note that to assign a category to a job, the title is analyzed using regular expressions–no fancy NLP or machine learning at this time–though we still review and can override the result. The automated classification has been extremely accurate thus far. However, as we expand the number of jobs and breadth of avenues by which they can be explored, we may have to do some actual ... data science. 😱
With the core categories in place, here’s how the jobs in 2020 broke out:
It’s worth noting how we find jobs–it’s not terribly scientific. Jobs are typically culled by scouring a variety of job boards and search engines, or revisiting the careers page of organizations we’ve already found. Once we’ve arrived at an organization’s careers page, we select all relevant jobs and leverage automation to load and classify them in our system. So while the path to add a job may initiate with a search for a particular type of role, e.g. Data Scientist, we always reach an organization’s careers page and look for roles in all categories while there. I’ll leave it as an exercise to the reader to interpret how representative these breakdowns may be of the data science industry at large.
In the new year, we’ve added an Other category to capture jobs in the realm of data science that don’t quite fit the core four. Some of the job titles we’ve included thus far are: Developer Advocate, Instructor, Mentor, and Data Journalist. Currently, the Other category comprises 3% of all jobs on the site.
I’ve talked about the broad categories that might demarcate boundaries within a data team, but what are we seeing in terms of specific job titles? Here are the most popular job titles on Outer Join in 2020.
At a glance, a few things jump out:
I hope to dive deeper on job titles in future posts–for example, I’d like to normalize these titles independent of seniority qualifiers to get a firmer sense of the distribution of core roles. As for the product, filtering jobs by title and experience level is on the roadmap for 2021. We also plan to be more proactive in locating junior roles and highlighting them on the site.
A list of organizations with counts of jobs that appeared on Outer Join in 2020 follows, but several caveats are in order:
For all of these reasons, please note that the numbers shown do not reflect the total number of data science roles that an organization listed in 2020. It’s simply a list of remote roles that we chose to publish on Outer Join.
As for who’s hiring in 2021: Browse our list of organizations that are currently hiring and filter by category as needed. You can also see who’s hiring for management roles and explore the aforementioned nonprofits.
I’ll wrap up here and get back to work so we have more to report down the line. I had a blast getting Outer Join underway in 2020 and our tiny team has been hard at work in 2021 to grow and improve the platform. If you have any feedback on this post or suggestions for Outer Join generally, we’d love to hear from you. Thanks so much for reading our inaugural blog post, I hope you come back for the SQL! 😇